Full AI software stack for driverless cars
Ambarella has developed a full software stack for autonomous driving (AD) using its low power AI chip. The company will demonstrate fully driverless car operation in January at the CES show.
The software stack was developed by the Vislab subsidiary in Italy and is primarily based on deep learning AI processing for all of its modular components, including environmental perception, sensor fusion and vehicle path planning. The stack is designed to run on the CVflow AI accelerators in Ambarella’s CV3-AD AI domain controller system-on-chip (SoC) family, minimizing power consumption and the processing load for its on-chip ARM cores.
The key to the low power performance is the use of standard definition maps that are generate din real time. This avoids the need for pre-generated high-definition (HD) maps and instead uses live environmental data from the vehicle’s sensing suite. This is more suited to handling difficult AD scenarios, such as downtown areas in large cities with roundabouts, narrow roads with parked vehicles, heavy traffic, construction, and a high density of pedestrians, cyclists and other vulnerable road users says Ambarella.
The chip also provides optimal processing for vision and radar sensing and fusion, and the software stack is modular to allow automotive OEMs to use any or all of the modules in combination with their own software intellectual property.
“This software stack draws on our more than 25 years of experience, building on the heritage of our VisLab Automotive R&D Team, whose AD stack has steadily evolved to take advantage of our latest AI SoCs,” said Fermi Wang, President and CEO of Ambarella. “Today, we are realizing the vision of our VisLab acquisition in 2015, which, in combination with our algorithm-first philosophy for designing purpose-built SoCs, has culminated in this platform that will enable automakers to deliver on the promise of autonomous driving.”
A key element of the stack is an AI-based planner. This dynamic vehicle path planning module enables a natural driving experience and is based on real world data from Ambarella’s evolving R&D vehicle fleet.
The latest members of the R&D vehicle fleet include a sensing suite that consists of mono and stereo cameras, as well as Ambarella’s Oculii 4D imaging radar; and all of the processing is being done by the CV3-AD.
The modular AD software and hardware provide OEMs with a scalable platform that enables them to select the CV3-AD family member with the right mix of performance and features for each model in their fleets, from ADAS assisted driving to autonomous features in mass-market vehicles.
In addition to its optimized AD software and hardware solutions, Ambarella provides automotive OEMs with a complete tool chain—both in-house and from ecosystem partners. This includes tailored tools for data collection (in conjunction with the CV3-AD’s on-chip encoder), simulation and annotation. Development is streamlined via accelerated software-in-the-loop (SIL) simulations using CV3-AD PCIe server cards, while Ambarella also provides a fully automated 3D annotation pipeline to support AI training.
The new Ambarella AD software stack will be demonstrated via fully autonomous test drives during CES in Las Vegas.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
